Project Summary/Abstract. While neuroinflammation is a proposed contributor to Alzheimer's disease (AD), the spatial and temporal relationship among inflammation, tau pathology and neurodegeneration remains unclear. Our long-term goal is to understand how microglial responses can be used to predict and monitor AD progression and potentially be manipulated for therapeutic purposes. Our objective is to determine how microglial activation, measured with PET imaging, is spatially and temporally related to tau pathology and neurodegeneration in AD. Our central hypothesis is that the microglial biomarker TSPO correlates with tau pathology and neurodegeneration cross-sectionally, and precedes the spread of tau and neurodegeneration longitudinally. We also postulate that the emerging CSF biomarkers YKL-40 and sTREM2 correlate with TSPO density, strengthening the argument that increased TSPO is a signal of pathological inflammation. Our rationale is that demonstrating how inflammatory mechanisms temporally correlate with tau pathology and neurodegeneration will identify time-points where immune-modulating therapeutics would be most beneficial. Inconsistencies in prior TSPO studies in AD are likely due in part to heterogeneity of age-of-onset, pattern of cognitive impairment, and topography of neurodegeneration in included patients. To overcome these confounds, we will enroll patients with specific clinical variants of AD ? amnestic AD, posterior cortical atrophy (PCA), and logopenic variant primary progressive aphasia (lvPPA). With this recruitment approach, we will acquire relatively uniform cohorts where disease is expected to have focal epicenters and spread in a predictable neuroanatomic distribution. We will acquire TSPO and tau PET, MRI markers of neurodegeneration, and CSF markers of inflammation in PCA, lvPPA, and amnestic AD patients, and in controls in a two-year longitudinal study. The specific aims are 1) Determine the topographical pattern of microglial activation in different AD subtypes, 2) Determine the spatial and temporal relationships between neuroinflammation and tau pathology in different AD subtypes, and 3) Determine the spatial and temporal relationship between neuroinflammation and MRI-based markers of neurodegeneration in different AD subtypes. For the first aim, we will measure TSPO using the state- of-the-art radioligand 11C-ER176, which has substantial advantages over earlier tracers and has been tested by our laboratory in AD patients. In the second aim, we will use the improved radioligand 18F-MK-6240 to measure tau pathology. In the third aim, we will use novel dimensionality reduction and multivariate statistical procedures to relate TSPO to tau pathology and MRI-based biomarkers of neurodegeneration in an unbiased, data-driven, and rigorous manner. Our innovative approach uses both PET and CSF measures of inflammation, clinically homogeneous cohorts, and novel statistical procedures to relate TSPO to tau pathology and neurodegeneration. The proposed research is significant, as it will inform use of TSPO PET to predict and monitor AD progression and response to novel therapeutics, and provide a robust imaging biomarker for future inflammatory targets.